538 Theme Day

#30DayChartChallenge Day 30

Day 30

Here we are at the final day of the #30DayChartChallenge! Today’s theme is 538 and browsing through the archives, I found this fun prediction they made for the 2015 NBA draft class. 538’s predictions weren’t very far off but I felt that the table that presented the results could be spruced up a little.

I also saw this as a great opportunity to use great_tables, a Python package that makes it easy to make beautiful HTML tables. For certain types of messages, a table might be more effective than a chart and it’s really cool to be able to elevate tables into proper data visualisations too!

import polars as pl
from great_tables import GT, md

# Data prep
draft = (pl.read_csv('https://raw.githubusercontent.com/fivethirtyeight/data/master/nba-draft-2015/historical_projections.csv')
    .insert_column(index=0, column=pl.Series(name="Rank", values=range(1, 1091)))
    .filter(pl.col('Draft Year')==2015)
    .filter(pl.col('Rank')<=16)
    .drop(['ID','Draft Year'])
)


# Great Table
table = (
    GT(draft)
    .tab_header(
        title = md("# Top College Prospects for the 2015 NBA Draft"),
        subtitle= "Ranked by 538's projected Statistical Plus/Minus (SPM)"
    )
    .tab_spanner(
        label="Probability of becoming", 
        columns=['Projected SPM',
        'Superstar',
        'Starter',
        'Role Player',
        'Bust']
    )
    .tab_source_note(md(
        '<br><div style="text-align: left;">'
        "**#30DayChartChallenge** | Day 30 | 538" 
        "<br> **Source**: https://fivethirtyeight.com/features/projecting-the-top-50-players-in-the-2015-nba-draft-class/" 
        "<br> **Made by**: www.ddanieltan.com"
        "</div>"
        "<br>"
    ))
    .fmt_number(columns='Projected SPM', decimals=2, pattern='+{x}')
    .fmt_percent(columns=['Superstar','Starter','Role Player','Bust'])
    .data_color(columns="Superstar",palette=["lightblue", "darkgreen"]
    )
    .data_color(columns="Bust", palette=["lightyellow","pink",]
    )
    .fmt_nanoplot(columns="Projected SPM", plot_type="bar")
)

table

Top College Prospects for the 2015 NBA Draft
Ranked by 538's projected Statistical Plus/Minus (SPM)
Rank	Player	Position	Probability of becoming
			Projected SPM	Superstar	Starter	Role Player	Bust
1	Karl-Anthony Towns	C		13.48%	42.72%	16.31%	27.50%
2	Justise Winslow	SF		8.35%	51.09%	17.68%	22.88%
3	Stanley Johnson	SF		6.78%	42.37%	27.85%	23.00%
4	Jahlil Okafor	C		5.87%	40.99%	23.55%	29.59%
5	D`Angelo Russell	PG		15.20%	34.23%	9.66%	40.91%
6	Dakari Johnson	C		2.13%	36.75%	41.76%	19.35%
7	Devin Booker	SG		7.34%	32.45%	39.02%	21.20%
8	Willie Cauley-Stein	C		4.71%	40.60%	24.32%	30.37%
9	Rondae Hollis-Jefferson	SF		1.46%	36.85%	39.25%	22.44%
10	Trey Lyles	PF		2.24%	35.13%	40.31%	22.32%
11	Sam Dekker	SF		0.87%	38.50%	32.91%	27.73%
12	Kelly Oubre Jr.	SF		1.98%	37.90%	33.80%	26.32%
13	Tyus Jones	PG		2.72%	41.46%	28.44%	27.38%
14	Kevon Looney	PF		0.83%	35.56%	37.49%	26.12%
15	Myles Turner	PF		2.94%	33.58%	31.85%	31.63%
16	Cliff Alexander	PF		0.76%	29.48%	40.66%	29.10%
#30DayChartChallenge | Day 30 | 538 Source: https://fivethirtyeight.com/features/projecting-the-top-50-players-in-the-2015-nba-draft-class/ Made by: www.ddanieltan.com

TIL

1. Polars has a very useful df.insert_column(index, column) that can specify the precise index to insert a new column at.

2. Great Tables provides md() method that can allow one to style any text using common markdown, this includes HTML tags.